Get all your news in one place.
100’s of premium titles.
One app.
Start reading
Top News
Top News

Openai Develops New Technique For AI Transparency Through Model Discussions

Amazon offers free credits for startups to use AI models including Anthropic

OpenAI has introduced a novel technique to enhance the transparency of AI models by facilitating communication between them. The company recently unveiled the research behind this innovative approach, which will be detailed in an upcoming paper. The fundamental concept involves engaging two AI models in a discussion to compel the more powerful one to elucidate its reasoning process, thereby aiding humans in comprehending how these models tackle problems.

OpenAI conducted tests by tasking AI models with solving basic math problems. The dominant model elucidated its problem-solving methodology, while the second model scrutinized the former's responses for inaccuracies. This technique is part of a series of recent releases by OpenAI that are integral to its mission of developing safe and beneficial artificial general intelligence.

The company also introduced a new scale to track its progress towards artificial general intelligence. This initiative comes on the heels of a period of upheaval in OpenAI's safety department, marked by the departure of key personnel. Concerns were raised about the company's commitment to safety following the exits of prominent figures involved in ensuring AI systems align with human interests.

Notable figures, including Tesla CEO Elon Musk and AI expert Stuart Russell, have expressed apprehensions about OpenAI's approach to AI development. Russell criticized the company's pursuit of artificial general intelligence without adequately validating safety measures as 'completely unacceptable.'

OpenAI's efforts to enhance transparency and safety in AI development are crucial steps towards building responsible and beneficial AI systems. The company's ongoing research and initiatives underscore its commitment to addressing the ethical and safety implications of artificial intelligence.

Sign up to read this article
Read news from 100’s of titles, curated specifically for you.
Already a member? Sign in here
Related Stories
Top stories on inkl right now
Our Picks
Fourteen days free
Download the app
One app. One membership.
100+ trusted global sources.